OpenAI exits and outage

OpenAI is dealing with two disruptions at once: another senior engineer has left, and ChatGPT just had a partial outage under investigation. (edtechinnovationhub.com) (status.openai.com) Ben Goodger, OpenAI’s head of engineering for ChatGPT Atlas, left for Google Labs and said on LinkedIn that he had “joined Google Labs” to build something new. EdTech Innovation Hub reported the move on April 22 and said Goodger had spent one year and 11 months at OpenAI after joining in June 2024. (edtechinnovationhub.com) Goodger previously spent 19 years at Google, where he helped create Chrome and ChromeOS, before moving to OpenAI. The same report said he will work at Google Labs with Brian Rakowski and Josh Woodward. (edtechinnovationhub.com) His exit follows three other April departures at OpenAI: Kevin Weil, vice president of OpenAI for Science; Bill Peebles, who led Sora; and Srinivas Narayanan, chief technology officer for business-to-business applications. CNBC reported those departures on April 17, alongside earlier role changes and leaves involving Fidji Simo, Kate Rouch, and Brad Lightcap. (cnbc.com) At the same time, OpenAI’s status page said the company was “currently experiencing issues” and was monitoring recovery from a ChatGPT Business problem after upgrades or added seats. The status page also lists aggregate uptime from January through April 2026 at 99.85% for ChatGPT and 99.99% for the application programming interface, or API. (status.openai.com) Separate outage reports on April 20 said OpenAI had acknowledged a broader partial outage affecting ChatGPT and Codex while it investigated the cause. Moneycontrol reported the incident affected 12 ChatGPT components and one Codex component, with users reporting login failures, gateway errors, and missing conversations. (moneycontrol.com) Those two developments land as OpenAI pushes deeper into enterprise sales. On April 8, OpenAI published a company post titled “The next phase of enterprise AI,” and on April 21 it published another titled “Scaling Codex to enterprises worldwide.” (openai.com) For customers, the issue is not just whether one outage gets fixed or one executive gets replaced. Companies that route internal search, coding, support, and document work through one model provider can be exposed when product leadership changes and shared services fail at the same time. (status.openai.com) (openai.com) OpenAI has not said that Goodger’s departure is connected to the outage, and CNBC reported the company described the science-team changes as a decentralization effort to move work closer to model, product, and infrastructure teams. The immediate facts are narrower: a key engineering leader left on April 22, and the company is still posting service incidents in April 2026. (cnbc.com) (edtechinnovationhub.com) (status.openai.com)